This R01 grant application requests 5 years of funding to extend the observation period of a cohort of first-time driving while impaired (DWI) offenders to 15 years. Subjects were first observed at the time of court-mandated referral for substance abuse evaluation to a screening program in Albuquerque, NM for a first DWI offense (1989-1992). A subset of 1,396 offenders was interviewed 5 years later where current and 12-month diagnoses of substance use and other psychiatric disorders were ascertained. A current secondary analysis grant examines the longitudinal database developed for the 5-year follow-up study. For the proposed study we will locate, and conduct structured, in-person interviews with the entire cohort of participants in the 5-year follow-up study. An estimated 71% of subjects can be located, with 5% deceased. A pilot study revealed a high (92%) willingness to participate, yielding an estimated total of 750 subjects (35% non-Hispanic White; 47% Hispanic; 15% American Indian; 3% other races; 56% of subjects are females). Interviews and questionnaires will measure the 15-year progression of alcohol use and other psychiatric disorders; examine attributes of subjects who achieved sobriety with and without the aid of formal and informal treatment; determine long-term rates of driving outcomes, including violations, crashes, and DWI arrests; and evaluate two alternative theoretical models (i.e., the Social-Control, and the Cognitive-Behavioral models) in predicting long-term, drinking/driving and traffic outcomes. Finally, through conduct of a qualitative study with a subset of offenders, we will assess subjects' self-report information and variability in long-term maintenance of sobriety and relapse. Prevalence of alcohol and drug abuse/dependence, and mental health diagnoses will be ascertained through well specified diagnostic criteria. Rates will be compared between genders and between Hispanic and non-Hispanic White subjects. Driving outcomes will be ascertained by self-report and using computer-based archival record data for those who still reside in New Mexico (estimated to be about 88% of subjects). Driving after drinking is the largest single cause of alcohol-related deaths in the U.S., and over one third of drivers in alcohol-related crashes have previous drunk-driving arrests. Yet few studies have examined long-term outcomes of convicted drunk drivers. Longer-term follow-up in this DWI sample, already described in rich detail, could add much to the body of knowledge regarding drunk drivers. The ability to collect long-term follow-up data coupled with theory development and an integrated empirical focus will provide a unique contribution to the development of prevention and treatment programs for both female and male offenders.